- the study was conducted on a new york subway over a 2 month period.
- the participants were all participants on the train weekdays between 11am and 3pm.
- a total of almost 4500 men and women
- slightly more white people than black people
- each trial lasted 7 and a half minutes. a team of 4 students boarded the train separately. two girls acted as observers, one boy was a confederate and the other a victim. one team had a black victim
- the drunk condition: the victim smells of alcohol and carries a bottle wrapped in a brown paper bag
- the cane condition: the victim appears sober and carries a cane.
- seventy seconds after the train pulls out the male victim staggers and collapses. if no help if offered the confederate steps in to help after either 70 seconds or 150 seconds.
- the observers recorded how long it took for help as well as information about the race, gender and location of all the passengers. the observers also noted comments overheard.
- the cane victim received spontaneous help 95% of the time but the drunk victim was spontaneously helped 50% of the time
- the cane victim was helped on average within 5 seconds but the drunk was helped after 109 seconds
- only 24% of the drunk victims were helped before the confederate stepped in but 91% of the cane victims were helped before the confederate stepped in
- the black victim received less help less quickly especially in the drunk condition
- there was a slight same race effect
- people in the immediate vicinity of the victim were more likely to help
- 80% of the first helpers were males
- an emergency situation creates a sense of arousal in a bystander
- this arousal is heightened if one feels a sense of identity with the victim or if one is physically close to the victim
- the longer the situation continues the greater the arousal
- arousal can be reduced by helping, going away or finding an excuse to why you cant help
- action depends on a cost vs benefit calculation
this study has high ecological validity as it is a field experiment in a normal setting, people also genuinely believed it was a real situation
this study has low research validity due to its lack of controls, it was conducted in as a field experiment thus extraneous variables couldn't be manipulated such as the occupation of then (if they were a nun they might be more likely to help) and age of participants. observers also noted difficulty effectively recording data during times when the carriage was busy.